Sunday, July 10, 2011

More Silliness from Claire Berlinski

I spent a little more time digging into the treasure trove of dreck that is Claire Berlinski's video oeuvre.

Ms. Berlinski, it seems, was present at a by-invitation only conference in Italy entitled "Great Expectations". It's hard to find anything about this conference online because, you see, it was "secret". But it's not hard to figure out the agenda. After all, the people present seem to have been

- Paul Nelson, creationist and remarkably unproductive philosopher for whom Paul Nelson Day was named. Watch Nelson squirm, evade, and do everything possible except answer the question of how old he thinks the earth is!

- Robert Marks, intelligent design proponent and writer of some remarkably silly papers about evolutionary algorithms

- David Berlinski, father of Ms. Berlinski, author of some remarkably bad popular books about mathematics, and contributor to such eminent scientific journals as Commentary. You can see Berlinski in all his superciliousness here. (Yet more superciliousness: David Berlinski on Gödel; David Berlinski on Popper.)

Berlinski claims we should be more open intellectually and some ideas are off limits to discussion. As usual, he's wrong. We just laugh at his ideas, and those of Nelson, because they are so incoherent. Even his daughter doesn't seem to buy it!

- Moshe Averick, creationist rabbi and sucker who apparently fell hook, line, and sinker for the scam that is "specified complexity", despite it having been debunked long ago

- Stephen Meyer, creationist, philosopher, and author of a a bad book containing misunderstandings of information theory. You can see his
videos here: Part 1A, Part 1B, Part 2, Part 2B, Part 3, and Part 4. It's funny to hear Meyer claiming that he "works on the origin of life". I wonder what experiments he has done and what labs he does them in. You can also hear Meyer extolling his creationist journal, Bio-Complexity, which has thus far published a grand total of 4 articles and one "critical review" -- every single one of which has at least one author listed on the editorial team page. It's a creationist circle jerk!

Meyer is allowed to repeat his bogus claim that "Whenever we find information, and we trace it back to its source ... we always come to an intelligence, to a mind, not a material process." Ms. Berlinski doesn't question him at all on this, despite the fact that it is evidently false.

- Richard von Sternberg, professional creationist martyr and co-author with Meyer of a drecky article filled with misunderstandings and misrepresentations.

- Michael Denton, author of a wildly wrong book, filled with misunderstandings about basic biology. Video here.

- perhaps Jonathan Wells. I can't be absolutely sure, but Meyer in this interview refers to cancer, and Wells is well-known for his wacky ID cancer theory. Of course, "journalist" Berlinski doesn't ask many hard questions. In the one hard question she does ask, about what are the best arguments against ID, Meyer can't even bring himself to mention the name of the person responsible.

You can watch Ms. Berlinski's "interviews" with Marks and Averick here (at a site where you have to pay them money to leave comments). You'd think with some of Marks' work on the record as being deficient, a journalist would have some hard questions to ask. But no, a giggling Ms. Berlinski lets Marks maunder on, making bogus claims like "All biological models of evolution which have been implemented in computer code only work because the information has been front-loaded into the program and the evolutionary process in itself creates no information" without asking any tough questions at all. (Marks, by the way, seems to think that Shannon coined the word "bit", when it fact it was Tukey.)

Reading the comments at that page is a real hoot, too. We have one commenter who "grew up with Information Theory from its early days", yet makes the false claims that (1) "there is still vigorous debate about which algorithms produce a truly random number; (2) "Whether you can determine the stopping point of a Turing machine is unsettled"; (3) "Many of these problems are essentially involved with extending Godel's Theorem beyond the realm of integers"; (4) "you have to consider what in Computation Theory is termed np-complete or in Penrose's term, non-computable". He also adds, helpfully, "I hope this sheds some light". Indeed it does, but not the kind of light he thinks.

It's just so funny to hear the people in Berlinski's interviews talk about how "orthodoxy" is "stifling" discussion when at least three of the attendees are members of conservative religious denominations that claim for themselves the right to determine truth for everyone else. Project much?

One thread that runs through many of Berlinski's interviews can be summarized as follows: "Waah! We're not taken seriously!" I'm not at all impressed with this. If you want to be taken seriously, don't hold "secret" conferences and make dark implications about being suppressed. If you want to be taken seriously, do some serious science; don't post videos with fart noises making fun of court decisions you don't like. If you want to be taken seriously, respond to critics in a professional way; don't depend on igorant attack-dog lawyers as your surrogates. If you want to be taken seriously, don't use credential inflation on your supporters and denigrate the actual scientific achievements of your detractors. You want some respect? Then earn it.

20 comments:

Anonymous said...

Well,

All the algorithms I have seen regarding random number generation are classified as pseudorandom number generators because they depend on a seed, like a timestamp for instance. So a given algorithm passed some statistical test, but this will only be to a certain level of confidence. The strength of these algorithms is that they are cheap to use. I used them to automate validation of power cycle times for hardware and software debug in design the stage for instance. Now hardware based random number generators produce data which can be viewed as more random. It is slower to use and so more expensive, and all hardware based generators are prone to bias and at some point in the hardware lifecycle the bias becomes pronounced enough that the data begins to fail the tests.

Patent a truly radom number generator and you stand a good chance of getting very wealthy.

Jeffrey Shallit said...

Dear Anonymous:

There is no "debate about which algorithms produce a truly random number" because anyone with a computer science education knows that deterministic algorithms don't generate random numbers. Read Knuth, The Art of Computer Programming, Volume 2, for more information.

Luckily, we aren't restricted to just deterministic algorithms. Physics, and in particular radioactive decay, give us a source of true randomness (at least according to current theory).

And, if you're happy with just pseudorandomness, not randomness, there are many interesting generators that have been developed by my thesis adviser, Manuel Blum, and others, that generate provable pseudorandomness under certain complexity-theoretic assumptions.

SLC said...

Dr. Berlinski, who has a legitimate degree in philosophy from Princeton, has passed himself off as a mathematician in the past. AFAIK, he has never published a math article in a peer reviewed math journal.

Jeffrey Shallit said...

SLC:

That's not quite true - but it is true that his record as a mathematician is quite poor. See my link the main article for a discussion of all of Berlinski's mathematical works.

http://recursed.blogspot.com/2008/04/david-berlinski-king-of-poseurs.html

SLC said...

Re Jeffrey

As Prof. Shallit states in the linked post, the three papers in question were largely philosophical papers. I am not familiar with mathematics journals but would Biomathematics be considered a mathematics journal (sounds more like an applied mathematics journal)?

Aside from the above considerations, would papers largely devoted to philosophical considerations qualify as a mathematics publications? As someone with a PhD in physics, I would not consider a submission to, say, Physics Today, which was largely philosophical in nature to be a physics publication.

Jeffrey Shallit said...

would Biomathematics be considered a mathematics journal (sounds more like an applied mathematics journal)

For my money, the definition of "mathematics" as "anything that gets reviewed in Math. Reviews" is probably pretty reasonable.

Anonymous said...

You clearly do not understand what bias means. Radioactive decay via what model kind sir? This method would be very expensive to use also. Get me proof, not a model.

Jeffrey Shallit said...

You clearly do not understand what bias means.

Based on what evidence? I'm a professional mathematician. Based on what you've written so far, you don't understand the issues so well.

Radioactive decay via what model kind sir?

Counting particle decays. Try reading here: http://www.fourmilab.ch/hotbits/

And it's easy to remove bias in not-quite-random bits; the basic ideas have been known for decades and date back to von Neumann.

This method would be very expensive to use also.

That's a different issue, entirely. Do you have a coherent criticism, or are you just flailing?

Anonymous said...

I recall being told 40 years agot hat we were taking the last few digits of a digital meter sampling the alternating current as a seed for a random number generator. I don't know if that's considered pseudorandom or random but it seemed like a good, inexpensive technique. Because the least significant digits were being used, it seems random enough.

David said...

I think Doug Axe was also present, I can't remember which video this was in, but one of the interviewees mentioned him giving a talk. So that makes 9/23 ID people.

Postbacc said...

Remarkably silly just about sums it up. I don't think they quite grasp the meaning of "peer reviewed"

CMarcusB said...

The animation from the Dover trial is absolutely priceless. Of course according to the link you provided, they are having a debate as to whether or not to include the fart noise. Dembski supposedly thinks it's sophisticated as he comments, "Calm yourselves everybody. An enhanced flatulent version is being worked on at this very moment. I will make it available. I do want to say this for the record, however. Many people regard the flatulent version as unsophisticated and even infantile. I want to suggest that in this postmodern age the flatulence in this animation actually serves as a sophisticated rhetorical device that mirrors the subtext of flatulence that runs throughout Judge Jones’s decision." They also wish to keep it for posterity when they "overturn" evolution, completely ignoring the fact that it does nothing but further reveal their childish inanity and disinterest in furthering scientific discussion, as rather it puts them on par with Beavis and Butthead. Not that we're offended though. As far as most of us are concerned they can just keep on keepin' on, while the adults continue the serious task of pursuing scientific inquiry and as a consequence bettering the human condition.

Anonymous said...

Radioactive decay is not a source of true randomness.

Science 101: we can model what is inside the box, but we can never know what is inside the box.

Thus you will never know by proof that any given model you champion will exhibit true randomness, especially in Physics.

So go ahead and delude yourself with those remaining so-so random bits you got so easily. Perhaps you can see your error and gain some integrity.

Why is true randomness so important to you?

Joshua said...

There is I think some (marginal) validity to the first of the four points. At this point there are as I understand it (I'm a number theory person not a complexity theory person) no provably pseudorandom number generators that don't reside on at least one conjecture, and that all the conjectures are things which are substantially stronger than P != NP (e.g. factoring being not in P). So in that context, 1 is sort of ok if you squint at it right.

The other three claims are so mindboggingly wrong that I suspect that the sort of almost correct nature of the first claim is more due to random chance more than anything else.

Joshua said...

Also, our anonymous friend's claim that it would be expensive to use a radioactive seed for random number generation is wrong. There are multiple actual implementations of this method. See for example http://www.fourmilab.ch/hotbits/

Making a radioactive number generator for oneself should also not be that hard. Bannanas are an excellent source of radioactivity (from all the potassium), and it is easy to make a Geiger counter. It shouldn't be that difficult to attach that to say an Arduino board and have that input a nice random stream into a computer. You'd have to replace the bannanas every few day but this might be a fun project for someone who likes to actually get their hands dirty and has a few free days.

The idea that such a system would be expensive seems to be based off the common misconception that radioactivity is rare. In fact it is quite common.

Jeffrey Shallit said...

Radioactive decay is not a source of true randomness.

And your proof of this is?

It's fine with me if you don't accept quantum mechanics, but I did say that my claim was dependent on that.

Why is true randomness so important to you?

It's not important at all - but the evidence suggests that it exists. If different evidence accumulates, I'll change my mind.

Why is the lack of randomness so important to you?

Anonymous said...

True randomness is not important to me either. I wanted to see if you really know the difference between facts and paradigms, since you talk a big game. You failed, but thanks for talking.

Jeffrey Shallit said...

Anonymous:

Everything in science is provisional, so the fact-paradigm distinction you seem to think is so important is more a matter of degree than of fundamental difference.

IdentifyYourself said...

To the person or persons calling themselves Anonymous: at least use a fake name.

KeithB said...

I am not sure how we got here, (Maybe anonymous is a pseudo random number generator!) but there is also LavaRnd:
http://www.lavarnd.org/what/index.html

And the classic SGI lavarnd that used Lava Lamps!
http://en.wikipedia.org/wiki/Lavarand